Full text of DOI:10.1111/j.1467-842X.2001.tb00586.x

Premature Death
Suicide prevention in Aboriginal communities:
application of community gatekeeper training
Abstract
Kim Capp, Frank P. Deane and Gordon Lambert
Illawarra Institute for Mental Health, University of Wollongong, New South Wales
Objective: Concern over the high rate of
suicide among Aboriginal people on the
south coast of NSW led to the development
of a project aimed at preventing youth
suicide in the Aboriginal communities of the
Shoalhaven. This paper describes the
development, implementation and
he Shoalhaven region has the symptoms who do not seek and receive ap-
seventh-highest population of propriate treatment or advice are at high risk
T
Aboriginal andTorres Strait Island- for suicide completion.6 In contrast, people
ers in New South Wales, representing 3.3% who receive appropriate help and advice
(2,491) of all Shoalhaven residents.1 During when suicidal are likely to reduce their level
the period from 1989 to 1996, the hospitali- of risk and distress.7 In the Shoalhaven the
sation rate in the Shoalhaven for attempted most significant negative predictor of a fur-
suicide among Aboriginal people was 9% ther suicide attempt for Aboriginal people
(61).2 The crude average annual hospitalisa- aged 15-29 years attending an emergency de-
tion rate for attempted suicide among partment for first attempt presentation was
Aboriginal people (502 per 100,000) is more having a referral for follow-up and the pa-
than four times higher than in the non- tient being seen by the mental health team.2
Aboriginal population (119 per 100,000) in Help seeking and facilitating appropriate
the Shoalhaven area. One of the most sig- help seeking has been shown to be an effec-
nificant positive predictors of a further sui- tive preventative strategy.
evaluation of the project.
Method: Following extensive consultation
with the Aboriginal community, a range of
culturally appropriate interventions were
developed. The main focus was a series of
community gatekeeper training workshops,
which aimed to increase the potential of
members of the Aboriginal community to
identify and support people at risk of
suicide and to facilitate their access to
helping services.
Results: Evaluation of the workshops
demonstrated an increase in participants’
knowledge about suicide, greater
cide attempt after an initial presentation is
In order to address these problems, an
being of Aboriginal or Torres Strait Islander early intervention program was undertaken
background. In the Shoalhaven the suicide in collaboration with the Shoalhaven Men-
rate forAboriginal people was approximately tal Health Service, University ofWollongong
22 per 100,000 population per annum, sig- and the Shoalhaven Aboriginal community.
nificantly higher than the non-Aboriginal In the initial stages, community consultation
population (13 per 100,000 per year) for the was conducted through a series of discus-
confidence in identification of people who
are suicidal, and high levels of intentions to
provide help. Attitudes, subjective norms
and barriers predicted intentions to help.
Conclusions: The project indicated
community members could be successfully
trained in the recognition of individuals at
risk of suicidal behaviour. Gatekeepers’
attitudes and perceived barriers to helping
predicted intentions to help those in need.
There is a need for longer-term follow-up to
assess the extent to which new knowledge
and skills are used in practice.
same period of time.2
sion groups8 to identify factors associated
It is self evident that the suicide rates are with suicide in the local area, barriers to help-
disproportionably high. Suicide is associated seeking, and how information on suicide
with depression and high levels of psycho- prevention could be disseminated within the
logical distress. However, few people seek Aboriginal community. Feedback from the
help when psychologically distressed. A discussion groups was used to inform the
recent Australian national survey of more design of a series of community gatekeeper
than 10,600 persons found that, while more training workshops. Of particular concern
than one in five adults meet the criteria for a were attitudes and beliefs suggesting it was
mental health disorder, 62% do not seek any not anyone else’s business if someone wanted
professional help.3 Evidence4 suggests the to kill themselves. Community gatekeeper
process leading to suicide is not an impul- training focuses on identifying key people
sive act but the end point in a pathological in the community who have contact with
regression, “a continuum beginning with youth and training them to recognise a per-
ideation, continuing to attempted suicide, son at risk of suicide, and to facilitate an
and ending with completed suicide”.5 appropriate referral to professional helping
Consequently, people displaying suicidal resources.9-11
Implications: Suicide awareness and skills
training have been demonstrated to be an
effective early intervention strategy.
Gatekeeper training empowers Aboriginal
communities and is generally accepted.
There is demand for such programs outside
the Shoalhaven. The project has a
methodological framework that can be
easily adapted by other communities.
(Aust N Z J Public Health 2001; 25: 315-21)
Correspondence to:
Submitted: February 2001
Revision requested: April 2001
Accepted: June 2001
Professor Frank Deane, Illawarra Institute for Mental Health, University of Wollongong,
NSW 2522. Fax: (02) 4221 5585; e-mail: frank_deane@uow.edu.au.
2001 VOL. 25 NO. 4
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
315

Capp, Deane and Lambert
Article
The Institute of Medicine (IOM)12 model identifies different
levels of suicide prevention according to the level of disorder/
distress in various populations targeted.13 It incorporates univer-
sal interventions that target whole populations at average risk,
selected interventions that target groups at increased risk, and
indicated interventions that target individuals at highest risk.
Community gatekeeper training has been classified according to
the IOM as “indicated/early intervention”.13
The second stage of the project, which is the focus of this
paper, involved the development and implementation of commu-
nity gatekeeper training. Eight free one-day workshops were held
in Nowra and the small, isolatedAboriginal communities ofWreck
Bay and Jerringa. Although other activities were carried out, this
was the main focus of the project evaluation.
The aim of the workshops was to increase the ability of the
local community to identify individuals at risk of suicide, mobi-
lise local informal helping networks and, where necessary, facili-
tate help-seeking behaviour. It became evident following
consultation with the community that in some cases the informal
community support networks were not well formulated or clearly
identifiable. The workshops endeavoured to provide clarification
of these resources and to also identify key individuals in the com-
munity who are willing to be a point of contact for people at risk
for suicide.
Although gatekeepers and gatekeeper training has been utilised
for the past 20 years as a strategy for identification and early
intervention in the prevention of suicide, a review of gatekeeper
programs in Australia found few were evaluated and those that
were usually focused on participant satisfaction surveys.9 Fur-
thermore, it appears that very few, if any, gatekeeper programs
have been implemented in Aboriginal communities.
The evaluation of the gatekeeper training in the present project
was an integral part of program planning and incorporated both
qualitative and quantitative measures to evaluate process and out-
come.The evaluation examines community attitudes towards sui-
cide and comparisons are made between Aboriginal participants
and non-Aboriginal groups from other studies. Participants were
also asked to rate a range of potential barriers to seeking or sup-
porting professional help seeking for those in need. Changes in
participants’ knowledge about suicide, confidence in identifying
people who are suicidal, intentions to seek help, and intentions to
refer those in need of help to the mental health service were
assessed. The evaluation also provides a preliminary test of the
utility of the Theory of Planned Behaviour14 to explain and pre-
dict intentions to help.
By increasing knowledge about suicide, increasing early iden-
tification of suicidal behaviours, establishing ways to help when
someone is suicidal, increasing individual confidence and will-
ingness to help when someone is identified as suicidal, clarifying
sources of help and how to successfully gain access to them (e.g.
overcoming potential barriers to help seeking, etc), it was felt
that the community would be better prepared and capable of help-
ing youth at risk of suicide.
The structure as well as the content of the workshop and the
resource manual15 were designed to include information identi-
fied by the discussion groups. This included local statistics, risk
factors for Aboriginal people and stressors within the Aboriginal
community associated with suicide.The components emphasised
at each workshop were determined by the specific need of the
group. For example, workshops targeting health workers included
information on legal issues such as duty of care, confidentiality
and assessment skills, whereas the general community required
training on awareness of suicide including warning signs, com-
munication and listening skills, how to respond to someone at
risk, and how to make an appropriate referral. Because the number
of participants was usually fewer than 10, this allowed for a vari-
ety of teaching methods including role-play and small group work.
An important component of the workshops was to have the co-
ordinator of the Shoalhaven Mental Health Service discuss the
service, address questions, and explain the role of the mental health
team and its place in suicide prevention and intervention. The
content of the workshops and the resource manual utilised both
recent published resources and information from the local dis-
cussion groups.
Method
Gatekeeper training program
The first stage of the project involved extensive consultation
with young Aboriginal people and local community groups. The
purpose was to obtain advice and information on appropriate sui-
cide prevention strategies for their community. Eleven discussion
groups were attended by 110 participants, of which 97 were from
the local Aboriginal community. To ensure information was
obtained from a broad cross-section of geographical, age and com-
munity representation, the following groups participated: Jerringa
Aboriginal Community; Nowra Department of Community Serv-
ices; Nowra Police; Nowra AboriginalYouth Touch Footie Team;
Aboriginal Education Assistants; Wreck Bay Aboriginal Com-
munity; Women’s Support Group; Oolong House (residential
alcohol and drug treatment centre); Dharawal Aboriginal Com-
munity; Elders Group; and Gerringong Aboriginal Touch Footie
Team. The information obtained from the discussion groups pro-
vided the basis upon which the project team developed the inter-
ventions and resources.8
The content covered in the workshops was delivered flexibly,
dependent on participants’ level of experience and knowledge.
The following information usually included: myths and facts; at-
titudes and beliefs; statistics (local and national); stressors (in-
cluding local perspective); risk assessment (including local
perspective); warning signs; communication skills; how to help
(including helping hand cards, and local contact numbers); per-
sonal and community support networks; referral strategies and
information; confidentiality; and duty of care.
These discussion groups were also crucial from a evaluation per-
spective as they provided informal feedback on the knowledge level
about suicide in the community, attitudes towards suicide, and the
barriers that prevented people from seeking or offering help.
316
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
2001 VOL. 25 NO. 4

Premature Death
Suicide prevention in Aboriginal communities
The workshops also introduced participants to the use of indi-
vidualised ‘survival maps’ that identified personal resources and
support networks of each participant.8
factorially robust in the SOQ.22 Of the original 16 items from this
subscale, six were selected based on either high eigen values
(>0.7), or being identified during discussion groups as strongly
help beliefs or attitudes (e.g. suicide is an acceptable means to
end incurable illness). Items from the SOQ were also selected
because this permitted the results to be compared with previously
collected Australian university student norms.17
Participants
Participants targeted were members of the Shoalhaven Abo-
riginal communities and workers from the Shoalhaven Aborigi-
nal Interagency Network (ie. Aboriginal health workers, youth
workers, Aboriginal education assistants). Community members
and workers were recruited by personal invitation, word of mouth,
hand-delivered fliers, and posters promoting the workshops at
community centres. The personal invitation was extended by the
ShoalhavenAboriginal suicide prevention/education officer at the
Aboriginal Interagency monthly meeting.
Knowledge
Knowledge about suicide was measured by 10 items based on
common misconceptions and facts about suicide. Eight of the
items were adapted from Kirk, Health Department of Western
Australia, and New South Wales Department of Health.23-25 Due
to concerns about the reading difficulty, minor modifications were
made to the questions to simplify the language (e.g. replacing
“Talking openly with individuals about suicide will increase the
risk of suicide” with “Talking about suicide will make people
want to do it”). The remaining two items consisted of misconcep-
tions that were repeated frequently by community members
attending the discussion groups (e.g. beliefs that mental health
workers were not trained in suicide prevention). All items had
‘true’, ‘false’ or ‘don’t know’ response options.
Participation was entirely voluntary. Those who attended the
workshops were invited to take part in the evaluation. Fifty-seven
people attended the workshops and fifty-three (93%) agreed to
participate in the evaluation. Although the workshops targeted
Aboriginal people, nine (16%) non-Aboriginal people attended
the workshops. Of the 48 Aboriginal people who attended, 44
(92%) consented to participate in the evaluation. For the purpose
of this evaluation, only the Aboriginal participants are included
in the analysis.The mean age of participants was 36 (SD=11 years)
and the range was 19 to 55 years. Ninety-one per cent (40) were
female and nine per cent (four) male.
Constructs pertaining to theTheory of Planned
Behaviour
In the Theory of Planned Behaviour,14 intentions are predicted
by attitudes, subjective norms and perceived behavioural control.
In turn, intentions are theorised to be the immediate precursor to
actual behaviour. Items for each of the measures were constructed
following procedures recommended by Ajzen and Fishbein.26
The need to limit the number of items resulted in assessment of
general components of attitude and subjective norms, but more
specific components of perceived behavioural control (PBC).
Perceived behavioural control incorporates barriers to behaviour
as part of its definition. For the purpose of the present project,
barriers were used to capture aspects of PBC. Barriers received
additional attention because of evaluative data obtained from the
discussion groups where concerns about a range of barriers were
identified by participants.
Of the participants, 18 (41%) were members of the general com-
munity, 20 (46%) were professionals including Aboriginal edu-
cation assistants, drug and alcohol workers, youth and health
workers, and six (13%) were TAFE students.
Measures
Suicide Opinion Questionnaire
Items from the Suicide Opinion Questionnaire15 were used to
measure aspects of attitudes towards suicide based on key issues
raised in the discussion groups. The SOQ has been used both in
Australia and internationally to examine attitudes within and be-
tween different groups and cultures.16-21 The SOQ consists of 100
items that cover a wide range of attitudes towards suicide.
For the purpose of this evaluation, only eight items were se-
lected from the standard 100 items. Community reluctance to
answer long questionnaires necessitated reducing the standard
measure. Six items were selected relating to the notion that sui-
cide is both “acceptable and normal”, and two questions on
‘lethality’, which assessed the attitude that suicide goes beyond a
mere ‘cry for help’. Items from this measure were included be-
cause of concern from the discussion groups that some members
may have viewed suicide as an acceptable means of resolving
problems. Others had expressed reluctance to help someone who
was suicidal because “it’s none of my business” or the belief that
“there was nothing you could do to help”. Each item was rated on
a five-point Likert scale from t ‘strongly agree’ to 1 ‘strongly
disagree’. The ‘acceptability and normality’ subscale is the most
Attitudes to help seeking were measured using four evaluative
semantic differential items with seven points between the anchors.
These were rated in response to the prompt, “For me to seek help
for someone else who is suicidal is:” (e.g. harmful to beneficial).
Scores on the four items were summed to provide a total attitude
score.26
Subjective norms were assessed using two items rated on seven-
point Likert scales. The first question, “Most people who are im-
portant to me would think I should seek help for someone who is
suicidal”, was rated from 1 ‘strongly disagree’to 7 ‘strongly agree’.
The response to the second question, “Generally, I want to do what
most people who are important to me think I should do”, was rated
from 1 ‘not at all’ to 7 ‘very much’. Responses to both questions
were multiplied to form the score for subjective norms.26
Barriers (PBC) were measured by six questions relating to a
2001 VOL. 25 NO. 4
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
317

Capp, Deane and Lambert
Article
Table 1: Comparison of Shoalhaven Aboriginal community sample and Australian university student sample on
‘acceptability and normality’ items from the Suicide Opinion Questionnaire (SOQ).
Shoalhaven Aboriginal community
Australian university students
(Domino, Niles & Raj, 1993)
Factor
Mean
SD
n
Mean
SD
n
Acceptability and normality (six items)
Notes:
2.31
0.74
44
2.96
0.64
82
The response scale assigns 5 ‘strongly agree’ to 1 ‘strongly disagree’.The lower the mean score the less acceptance of suicide.
variety of internal and external control factors. Four items were
developed from the discussion groups that assessed the possible
problems or obstacles that prevented participants from seeking or
offering help for someone who was suicidal. They related to
issues of confidentiality, trust and privacy (e.g. “Kooris keep things
to themselves rather than talking about it.”). A further two non-
specific barrier items were included (e.g. “There is nothing to
stop me from obtaining help for someone who is suicidal”). All
items were rated on a Likert scale from 1 ‘strongly disagree’ to 7
‘strongly agree’. Items were summed to form a total barrier score.
Intentions were measured by two questions. The first was a
broad measure of ‘intentions to help’ and the second measured
‘intentions to refer to the mental health service’. The score from
the former question provided a general measure of help seeking
intentions for testing theTheory of Planned Behaviour constructs.
The questions were measured on a Likert scale with participants
rating their responses from 1 ‘extremely unlikely’to 7 ‘extremely
likely’. These items followed the same structure as those used in
previous research assessing help seeking intentions.27-29
put on overhead transparencies and read out to participants. The
language and cultural appropriateness was assessed by the Abo-
riginal members of the project team, followed by the Shoalhaven
Aboriginal Mental Health Advisory Group.
The research received ethical review and approval from the
Wollongong University Human Ethics Committee.
Results
Sample sizes vary for some analyses due to small amounts of
missing data on some variables.
Suicide Opinion Questionnaire
Table 1 shows the means and standard deviations for compari-
son between Aboriginal and university student samples. The
results indicate there is very little difference between groups and
both generally view suicide as not acceptable or normal behaviour.
Repeat measures of knowledge, intentions and
confidence
Paired sample t-tests were performed to assess whether change
occurred in areas of knowledge, intentions to help, intentions to
refer to the mental health service, and confidence in identifying a
suicidal person. Table 2 presents the results.
Confidence to identify a person at-risk for suicide was meas-
ured by one item and scored on a Likert scale from 1 ‘not at all
confident’ to 7 ‘very confident’.
Procedure
Development of questionnaires
The results on knowledge show a significant increase in
the number of true/false items answered in the keyed direction
from Time 1 to Time 2 (i.e. over the course of the workshop),
t(40)=-7.05, p<0.001 (two-tailed, n=41). For confidence, there
was a significant increase over time, t(39)=-1.77, p<0.05 (two-
tailed, n=39). There was no significant change in intentions to
help from Time 1 to Time 2. In contrast, the likelihood ratings of
“intentions to refer to the mental health service” decreased sig-
nificantly fromTime 1 to Time 2, t(39)=2.11, p<0.05 (two-tailed,
n=40). A reduction in intentions to refer to the mental health cen-
tre was an unexpected finding, therefore an attempt was made to
explore what may have contributed to this finding by conducting
post hoc analyses.
The content of the questionnaires were reviewed and devel-
oped in consultation with the Aboriginal suicide education/pre-
vention officer and the Shoalhaven Aboriginal Advisory Group.
In planning the evaluation, it was necessary to incorporate feed-
back indicating that Aboriginal communities were “sick of doing
long questionnaires”. There were reports of a number of exam-
ples where researchers had simply handed out questionnaires that
were subsequently not completed. Thus, a very conservative
approach was taken in order to minimise demand on participants.
In any single administration block the number of items was re-
stricted to 10 or less. Some components were assessed once (SOQ,
TPB variables) and others were repeated (knowledge, intentions
and confidence) to evaluate change. To ensure a seamless evalu-
ation, the items were integrated into the goals of the workshop
and administered in a sequential manner. Used in this way, the
questions stimulated group discussion and were seen as relevant
to the participants. The questionnaires were administered by a
member of community who was trained and funded by the project.
To allow for any problems with literacy, the questionnaires were
Self-determination has been a strong political agenda for many
Aboriginal communities.30,31 We speculated that as the confidence
of members of the Aboriginal community increased, so too did
their beliefs that they may be more able to help people who were
suicidal within their own community networks. Thus, they may
be less likely to refer to the mental health centre because they are
more confident of managing those individuals themselves.
Support for this hypothesis would be present if changes in confi-
318
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
2001 VOL. 25 NO. 4

Premature Death
Suicide prevention in Aboriginal communities
Table 2: Means and standard deviations for comparisons between repeat measures.
Time 1
Time 2
Measures
N
M
SD
M
SD
Knowledge
41
40
40
39
5.59
6.58
6.50
4.87
1.84
0.78
0.91
1.88
7.59
6.50
6.13
5.44
1.79b
Intentions to help
0.85
Intentions to refer to MHS
Confidence to ID suicide risk
1.24a
1.57a
Notes:
(a) p<0.05
(b) p<0.001
Knowledge scale maximum score = 10; Intentions scales ranged from 1 ‘Extremely unlikely’ to 7 ‘Extremely likely’; Confidence scale ranged from 1 ‘Not at all confident’
to 7 ‘Very confident’.
dence, specifically an increase in confidence represented by the
difference between Time 1 and Time 2, were negatively corre-
lated with changes in intentions to refer to the mental health cen-
tre. Similarly, there should be a positive correlation between
increased confidence and intentions to help.
Barriers generated from the discussion groups concerning
issues of trust, privacy and confidentiality were perceived as be-
ing major obstacles preventing people from seeking or offering
help. A detailed description of the barriers is provided in the re-
port to the NSW Centre for Mental Health.32
To explore the relationship between intentions and confidence,
Pearsons correlation coefficients were calculated using the mean
difference score (Time 1 to Time 2) for the confidence and inten-
tion measures. The results show a negative but non-significant
correlation (r=-0.21) between changes in confidence and inten-
tions to refer to the mental health service. In contrast, there was a
positive correlation (r=0.13) between increased confidence and
increased intentions to help (also not significant).
Discussion
The results from the SOQ suggested that in theAboriginal sam-
ple, suicide was not considered normal or acceptable behaviour.
This is not surprising, as the people who attended the workshop
were self-selected and highly motivated to prevent suicide in their
community. At a descriptive level, the acceptability of suicide
does not appear any greater in this Aboriginal sample when com-
pared with an Australian university sample.18 However, consider-
able caution is needed before attempting to generalise these
findings to the wider Shoalhaven Aboriginal community. An
examination of the discussion group transcripts painted a very
different picture of attitudes.The following comments came from
discussions with young Aboriginal men: “You know that’s their
business. If you want to die that is their business, it is their life”,
“... he just hung there. I just sat there right through it”.
Predicting intentions
To determine the predictors of help seeking intentions, stand-
ard multiple regression analyses was completed. Table 3 provides
the correlation matrix and regression coefficients for predicting
intentions to help someone who is suicidal.The likelihood of seek-
ing help for somebody who was suicidal was the dependent vari-
able, with attitudes, subjective norms, and barriers as the
independent variables. These variables produced a significant re-
gression coefficient and together accounted for 37% of the vari-
ance in intentions ratings, F(3,38)=7.29, p=0.001. Both attitudes
(standardised Beta=0.43, t=3.01, p=0.005) and barriers (stand-
ardised Beta=-0.28, t= -2.05, p<0.05) were unique significant
predictors of help seeking intentions. Positive attitudes to seek-
ing help were associated with higher help seeking intentions.
Barriers were negatively related to intentions.As perceived barri-
ers increased, intentions to help decreased.
It is reasonable to say that there is a diverse range of attitudes
related to suicide among members of theAboriginal community.There
is a need for a more detailed and comprehensive examination of a
broad cross-section of the Aboriginal community to obtain a more
complete assessment of attitudes. There is also a lack of data on
differing cultural constructions and understandings of suicide and
future research should more comprehensively address these.33
There was a significant increase in knowledge over the course
Table 3: Standard multiple regression of predictor variables on intentions to help.
Variables
2
3
4
M
SD
Beta
0.43b
1. Intentions to help
2. Attitudes
0.54b
0.22
0.29a
-0.36b
-0.25
0.18
6.55
6.33
0.80
1.13
3. Subjective norms
4. Barriers/PBC
32.88
3.47
15.73
1.10
0.15
-0.28a
Notes:
(a) p<0.05.
(b) p<0.01.
r2=0.37, adjusted r2=0.32, r=0.60, n=42
Listwise deletion was used resulting in some missing data for this analysis.
2001 VOL. 25 NO. 4
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
319

Capp, Deane and Lambert
Article
of the workshops. The finding that knowledge increased was
similar to the outcomes from other gatekeeper training pro-
grams.34-36 However, the increase in knowledge may only be short
term and long-term follow-up would be necessary to examine
whether this information is retained.
negative correlation (r=-0.21) between changes in confidence and
intentions to refer to the mental health centre. As confidence in-
creased, intentions to refer to the mental health centre decreased.
These findings are highly speculative because the correlation co-
efficients were not significant. However, they are intuitively con-
sistent and are presented as a cautionary consideration for those
using such training in the future.
It is also relevant to comment on two statements where the cor-
rectness of the keyed response could be questioned. For example,
the item, “Mental health workers are trained in helping people
who are suicidal” (keyed, “true”) is arguable. Bognar and
Harmatz37 reviewed the clinical psychology programs in the
United States and found only 40% offered formal training in the
study of suicide. It was evident from the discussion groups that
people did not think mental health workers were trained in sui-
cide prevention. It is important that the community is made aware
of the various roles of mental health workers as this is often the
only crisis service available. The second statement concerns the
debate over whether suicide is a mental health issue or not. This
issue had been raised during the discussion groups. For example,
one participant made the following comment, “Well, you’re not
always mental if you try to suicide are you?” Once again this is
debated in the literature with some suicide prevention programs
deliberately de-emphasising the link between mental illness and
suicide. Garland and Zigler38 suggested that attempting to
destigmatise suicide can result in “normalising the behaviour and
reducing potentially protective taboos” (p. 174). Other groups
argue suicide is a reflection of social issues.
A further explanation for these findings requires an understand-
ing of the history and relationship between the Aboriginal com-
munity and mental health services. From the discussion groups,
it was evident people had a reluctance to use the mental health
service.There was a perception that non-Aboriginal mental health
workers would not understand the experiences of an Aboriginal
person. Furthermore, the community was suspicious of mental
health services. Combined with a strong political agenda of self-
determination it is possible that training may empower Aborigi-
nal people to use their own available helping resources, rather
than calling for outside help.
The Theory of Planned Behaviour14 was used to help deter-
mine the factors most important in the prediction of people’s will-
ingness to help.Attitudes, subjective norms and barriers explained
37% of the variability in intentions. This finding is similar to
prior studies examining help seeking intentions.27,28 Both attitudes
and barriers were unique predictors of intentions.
The discussion groups and workshops provided an ideal
opportunity to obtain qualitative data regarding attitudes and bar-
riers. Barriers concerning privacy, confidentiality and trust were
described as major obstacles preventing people from help seek-
ing. Privacy issues concerned a blurring of professional and per-
sonal roles associated with small communities. It was not clear
whether there is a general reluctance to discuss personal prob-
lems within the community or whether it was specific to suicide.
It was reported that some Shoalhaven Aboriginal families have
experienced suicide of more than one family member and have
refused help from outside agencies. The workshops aimed to
assist in breaking this cycle of avoidance and to facilitate help
seeking. One participant said, “If you can influence one person in
the family they can open the door [to help]”.
Both these statements generated helpful discussion in the work-
shops, but it was important to clearly identify local mental health
services as appropriate sources of help for suicidal individuals.
There was also a significant increase in confidence in identify-
ing individuals who were suicidal. Increases in knowledge and
confidence do not guarantee improved skills. Future research
should begin assessing skill acquisition. One possible measure in
the future would be Neimeyer and MacInnes39 25-item Suicide
Intervention Response Inventory.This would better show whether
skills training is effective when dealing sensitively with suicidal
persons.
There was no increase in intentions to help someone display-
ing suicidal behaviour, but these intentions remained extremely
high both atTime 1 (M=6.58, out of a maximum 7) and atTime 2
(M=6.50). Such high baseline levels of helping intentions were
most likely a function of a highly motivated sample. This finding
is consistent with other gatekeeper programs, which did not show
expected gains in knowledge, and this was attributed to partici-
pants’ high levels of pre-existing knowledge.40
On the other hand, it may be that they do not have the tradi-
tional source of help wisdom, or structure previously provided by
elders. For example, one participant made the following com-
ment, “Previously, elders were there to go to, [people] took
notice of them, [it] still does take place, but not enough”. One of
the participants has been referred to as an “elder” but she was
reluctant to accept this role, because she felt “too young”. An-
other important consideration is that Aboriginal people are not
getting “old”. High death rates in younger and middle Aboriginal
adult life, with a life span some 20 years shorter compared with
non-Aboriginal Australians,41 has resulted in fewer elders.
It may be that further research is needed to help identify which
specific attitudes and barriers have the strongest influence on pre-
venting people from seeking help for someone who is suicidal.
The Theory of Planned Behaviour provides a promising start to
this process and an additional step will be to examine the extent
The likelihood rating of “intentions to refer to the mental health
service” decreased significantly from Time 1 to Time 2. This was
an unexpected finding but may also be attributed to very high
levels of baseline intentions and possibly regression to the mean.
However, to further explore this finding the relationship
between changes in intentions (“intentions to help”, and “inten-
tions to refer to MHS”) and confidence was examined. The
results showed a positive relationship (r=0.13) between changes
in confidence and intentions to help. In contrast, there was a
320
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
2001 VOL. 25 NO. 4

Premature Death
Suicide prevention in Aboriginal communities
12. Mrazek PJ, Haggerty RJ, editors. Reducing the Risks for Mental Disorders:
Frontiers for Preventive Intervention Research. Washington(DC): National
Academy Press, 1994.
13. Silverman MM, Felner RD. The place of suicide prevention in the spectrum
of intervention: Definitions of critical terms and constructs. In: Silverman
MM, Maris RW, editors. Suicide Prevention: Toward theYear 2000. NewYork:
Guilford Press, 1995.
14. Ajzen I. The theory of planned behaviour. Organ Behav Hum Decis Process
1999;50:179-211.
15. Curtis J, Lambert G, Capp K, Deane FP, de Ramirez D, Boota K. Shoalhaven
Aboriginal Suicide Prevention Project: Community Gatekeeper Training,
Participant Resource Manual. Wollongong(NSW): Illawarra Institute for
Mental Health, University of Wollongong, 2000.
to which intentions predict actual helping behaviours.
Furthermore, there is a need to evaluate the long-term effec-
tiveness of sustained community gatekeeper training in Aborigi-
nal communities. For example, the evaluation process should
assess how many participants identified and tried to facilitate help
seeking for someone who was at risk of suicide. Ideally, such
evaluation should include whether there are increases in referral
rates and engagement with mental health services or other help-
ing resources (e.g. elders, community members,Aboriginal Medi-
cal Services).
16. Domino G, Moore D, Westlake L, Gibson L. Attitudes toward suicide: A fac-
tor analytic approach. J Clin Psychol 1982;38:257-62.
17. Domino G, Niles S, Raj SD. Attitudes toward suicide: a cross-cultural com-
parison of Singaporean and Australian university students. Omega 1993-
1994;28:125-37.
Acknowledgements
18. Domino G, Su S. Conservatism and attitudes toward suicide: A study of Tai-
wanese-American and U.S. adults. J Death Dying 1994-1995;30:(2):131-43.
19. Leane W, Shute R. Youth suicide: The knowledge and attitudes of Australian
teachers and clergy. Suicide Life Threat Behav 1998;28:(2):165-73.
20. Leenaair A, Domino G. A comparison of community attitudes towards sui-
cide in Windsor and Los Angeles. Can J Behav Sci 1993;25:(2):253-66.
21. Lester D, Akande A. Attitudes about suicide among the Yoruba of Nigeria. J
Soc Psychol 1994;134:(6):851-3.
22. Domino G.Test-retest reliability of the Suicide Opinion Questionnaire. Psychol
Rep 1996;78:1009-10.
23. Kirk WG. Adolescent Suicide: A School-Based Approach to Assessment and
Intervention. Illinois: Research Press, 1993.
24. Health Department of Western Australia. Making a Difference: Youth Suicide
Prevention Manual. Perth: Print Image, 1994.
25. Centre for Mental Health. New South Wales Youth Suicide Prevention Train-
ing Manual. Sydney: NSW Department of Health, 1998.
26. Ajzen I, Fishbein M. Understanding Attitudes and Predicting Social Behav-
iour. New York: Prentice-Hall, 1980.
27. Deane FP, Todd DM. Attitudes and intentions to seek professional psycho-
logical help for personal problems or suicidal thinking. J Coll Stud Psychother
1996;10:45-59.
28. Deane FP, Skogstad P, Williams MW. Impact ofAttitudes, Ethnicity and Quality
of Prior Therapy on New Zealand Male Prisoners’ Intentions to Seek Profes-
sional Psychological Help. Int J Adv Couns 1999;21:55-67.
29. Carlton PA, Deane FP. Impact of attitudes and suicidal ideation on adoles-
cents intentions to seek professional psychological help. J Adolesc 2000;
23:35-45.
The project was funded from the Commonwealth Funding for
Rural Youth counselling under the National Youth Suicide Pre-
vention Strategy. The project could not have been undertaken
without the advice, support and endorsement of the Shoalhaven
Aboriginal Mental Health Advisory Group, the Aboriginal Medi-
cal Service, local Aboriginal organisations, and members of the
participating communities of Jerringa, Wreck Bay, and Dharawal.
Appreciation is also extended to the Aboriginal and non-
Aboriginal health workers at the Nowra Community Health
Centre with special thanks to Iris McLeod, coordinator of
Aboriginal Health. The contribution of the project team members
is acknowledged, in particular Dawn de Ramirez, Kevin Boota
and Joanne Connolly. Dawn de Ramirez, the Shoalhaven
Aboriginal suicide prevention/education officer, has endorsed this
report as an accurate representation of the activities of the project,
and the Shoalhaven Aboriginal Mental Health Advisory Group
supports the publication of this information.
30. Royal Commission Into Aboriginal Deaths In Custody. Report. Canberra:
AGPS, 1991.
31. Swan P, Raphael B. Ways forward. National Consultancy Report on Aborigi-
nal and Torres Strait Islander Mental Health: Part 2. Canberra: Australian
Government Printing Service, 1995.
32. Capp K, Deane FP, Lambert G. Shoalhaven Aboriginal Suicide Prevention
Project. Report to NSW Health Department. Wollongong(NSW): Illawarra
Institute for Mental Health, University of Wollongong, 2000.
33. Graham A, Reser J, Scuderi C, Zubrick S, Smith M, Turley B. Suicide: An
Australian Psychological Society discussion paper. Aust Psychol 2000;
35(1):1-28.
34. Kalafat J, Elias M. Suicide prevention in educational context: broad and nar-
row foci. In: Silverman MM, Maris RW, editors. Suicide Prevention: Toward
The Year 2000. New York: Guilford Press, 1995.
35. King R. Suicide prevention: dilemmas and some solutions. Rural Soc
1994;4:(3):2-6.
36. Shaffer D, Garland A, Gould M, Fisher P, Trautman P. Preventing teenage
suicide: A critical review. J Am Acad Child Adolesc Psychiatry 1988;
27:(6):675-87.
37. Bognar B, Harmatz M. Clinical psychology graduate education in the study
of suicide: Availability, resources, and importance. Suicide Life Threat Behav
1991;21:70-5.
38. Garland AF, Zigler E. Adolescent suicide prevention. Am Psychol
1993;48:(2):169-82.
References
1. Australian Bureau of Statistics. Census (unadjusted): Aboriginal Profile in
Statistical Local Areas, New South Wales. Canberra: ABS, 1996.
2. Preston W, Morrell S, Blackmore K, Elkington J. Best Practice in Suicide
Management: Combined Services Approach to Management of Suicidal Pres-
entations to An Emergency Department. Wollongong(Aust): Illawarra Area
Health Service, Shoalhaven District Memorial Hospital, 1998.
3. Andrews G, Hall W, Teeson M, Henderson S. The mental health of Austral-
ians. National Survey of Mental Health and Wellbeing Report 2. Canberra:
Commonwealth of Australia, 1999.
4. Novick J. Attempted suicide in adolescence: the suicide sequence.
In: Maltsberger JT, Goldblatt MJ, et al., editors. Essential Papers of Suicide.
Essential Papers in Psychoanalysis. New York: New York University Press,
1996. p. 524-48.
5. Cole DE, Protinsky HO, Cross LH. An empirical investigation of adolescent
suicidal ideation. Adolescence 1992;27(108):813-8.
6. Rosenberg ML, Eddy DM, Wolpert RC, Broumas EP. Developing strategies
to prevent youth suicide. In: Pfeffer CR, editor. Suicide Among Youth: Per-
spectives on Risk and Prevention. USA: American Psychiatric Press, 1989.
7. Rudd MD, Rajab MH, Orman DT, Stulman DA, Joiner T, Dixon W. Effective-
ness of an outpatient intervention targeting suicidal young adults: prelimi-
nary results. J Consult Clin Psychol 1996;6:179-90.
8. Capp K, de Ramirez D, Boota K. Suicide is everybody’s business — building
community resilience. Paper. Presented at the NSW Aboriginal Mental Health
Conference ‘Moving Forward Together’; 1999; Sydney, Australia. In press.
9. Frederico M, Davis C. Gatekeeper Training and Youth Suicide Prevention.
Report for Youth Suicide Prevention Initiative: Education and Training Con-
sultancy. Canberra: Department of Health and Family Services, 1996.
10. Mitchell P. Early intervention: Its place in the prevention of youth suicide.
Aust Early Interv Newsl 1998;5:2-7.
39. Neimeyer R, Pfeiffer AM. The ten most common errors of suicide interven-
tionists. In: Leenaars AA, Maltsberger JT, Neimeyer RA, editors. Treatment
of Suicidal People. Washington: Taylor Francis Publishers, 1994.
40. Beckman LJ, Mays VM. Educating community gatekeepers about alcohol
abuse in women: Changing attitudes, knowledge and referral practices. J Drug
Educ 1985;15:289-309.
41. Bhatia K, Anderson P. An overview of Aboriginal and Torres Strait Islander
Health: Present Status and Future Trends. Canberra: Australian Institute of
Health and Welfare, 1995.
11. Potter LB, Powell KE, Kachur SP. Suicide prevention from a public health
perspective. In: Silverman MM, Maris RW, editors. Suicide Prevention: To-
ward the Year 2000. New York: Guilford Press, 1995.
2001 VOL. 25 NO. 4
AUSTRALIAN AND NEW ZEALAND JOURNAL OF PUBLIC HEALTH
321